IMPACT OF DISAGGREGATED PUBLIC EXPENDITURE ......Harvey Leibenstein (1957) posited that developing...

International Journal of Development and Economic Sustainability

Vol.7, No.3, pp. 64-79, May 2019

___Published by European Centre for Research Training and Development UK (www.eajournals.org)

64 Print ISSN: 2053-2199 (Print), Online ISSN: 2053-2202(Online)

IMPACT OF DISAGGREGATED PUBLIC EXPENDITURE ON ECONOMIC

GROWTH OF SELECTED AFRICAN COUNTRIES: A PANEL VECM

Favour C. Onuoha, Moses Oyeyemi, Agbede

Economics Department, Evangel University Akaeze, P.M.B. 129 Abakaliki Km 48 Enugu

Abakaliki Expressway, Okpoto Ebonyi State, Nigeria.

___________________________________________________________________________

ABSTRACT: The study investigated the long-run and short-run equilibrium relationship

between economic growth and disaggregated public expenditure in selected West African

Countries with panel data spanning 1990-2017. The study employed panel co-integration

based on Pedroni and Panel Vector Error Correction Model (PVECM) with Engle and

Granger´s procedure for empirical analysis. The findings revealed that expenditure on

infrastructure, health and education have positive impact on economic growth at about 2%,

6% and 2% respectively, but only expenditure on infrastructure is significant. Defence

expenditures and education expenditures at both lags have indirect and insignificant influence

on economic growth while health expenditure has direct and insignificant impact on economic

growth at all lags. The study recommends policy makers to focus on developing health,

infrastructure and education sectors which has not contributed significantly enough to

economic growth in the selected African countries.

KEYWORDS: Health, Education, Infrastructure, Defence, Economic growth, PVECM

INTRODUCTION

Background to the study

Over the years, public expenditure has been increasing in geometric term through various

government activities and interactions with its Ministries, Departments and Agencies

(MDA’s), Niloy, Emranul and Denise (2003). This sharp rise in public expenditure in African

countries caught researchers’ attention for more investigation owned to the poor welfare of the

citizens, despite the gross government expenditure spree for decades. There has been growing

concern about the extent to which government expenditure has impacted the economic growth

in African countries. The rising cost of governance remained a challenge by African countries;

the public expenditure size has expanded which has generated interest in both developed and

developing world to optimise the size of government. The need to provide and expand the

tentacles of public goods becoming too obvious and unavoidable recognised, mismanagement

and misappropriation of public expenditure in the economy cannot be underestimated, coupled

with the pressing demand to expand and cater for the rising population.

Amidst the unresolved foregoing controversies, most African countries are still faced with

monumental development problems, therefore, the policy makers emphasized on the roles of

public sector expenditure as important instrument which the government can apply to restore

some economic problems such as reduction in inequality, inflation, fall in exchange rate,

unemployment, dwindling oil price and the desire to restore the economy on the part of full

employment, price stability, balance of payment equilibrium and above all, increase in

economic growth. However, it has been argued that, the rising state of public sector expenditure

http://www.eajournals.org/


Vol.7, No.3, pp. 64-79, May 2019



contributed to economic growth, this has continued to generate series of debate among scholars,

the empirical and theoretical positions on the subject is quite diverse and still remain mixed.

According to lyoha and Oriakhi (2002), the overviews of some African countries performance

compared to the developed world are not comparable, when the world's economy grew at an

annual rate of close to 2% from 1960 to 2002, Africa growth performance was in dismal, from

1974 through the mid-1990s, growth was negative, reaching -1.5% in 1990 to 1994. Oteng-

abayie (2011) noted that one half of the African continent inhabitants live below the poverty

line which is still the same till date. There has been a continuous decline of per capita GDP in

most African countries; social indicators among the worst in the world, infant mortality rate

recorded highest, life expectancy at low ebb with many problems bedevilled the African

nations.

Therefore, it is against these issues raised above that this study examine whether gross public

expenditure has any impact on economic growth (proxied by gross domestic product) in

selected African countries has become necessary. Hence, the study provides answers to the

impact of public expenditure of selected African countries on the economic growth. The study

is structured to the following arrangement, section one captures the background to the study,

section two focuses on detailed theoretical propositions and empirical review. Section three

explains the method adopts to analyse the data while section four shows outcome results and

interpretations. Finally, section five entails summary, conclusion and policy recommendations.

LITERATURE REVIEW/ THEORETICAL UNDERPINNING

Wagner (1893), Peacock and Wiseman (1961) and others great economists have formulated

different theories on public expenditure and economic growth. Wagner viewed public sector

expenditure as a behavioural variable that positively dictates if an economy is growing.

However, the neo classical growth model developed by Solow(1956) opined that the fiscal

policy does not have any effect on the growth of national output. In the case of public

economist, most of the studies opined that public expenditure is majorly to increase economic

growth which is in tandem with Wagner’s hypothesis. Solow (1956) argued that invention

through fiscal policy helps to improve failure that might arise from the inefficiencies of the

market. Similarly, Dar and Amir (2002) postulated that in the endogenous growth models,

fiscal policy is very crucial in predicting future economic growth. Barro (1990), Barro and Sali-

i-Martins (1992) and Roux (1994) all noted that the expansion of government expenditure

contributes positively to economic growth.

Rostow (1960) attests the need for government intervention through massive injection of

capital into economy to hasten massive development process in developing countries to

stimulate the growth of the economy. Similarly, the big push theory encourages huge capital

injection to drive the economy growth from the slow and epileptic state. The theory opined that

if a low level of equilibrium trap exists, there is an urgent need for a critical minimum effort

required to escape from economy stagnancy due to low saving and low income that may persist.

It is noteworthy that Harrod (1960) and Domar (1946)economic growth model hold that, an

impoverished individual have meagre or no saving, government intervention for massive

capital injection in such an economy is to improve the people condition to move from abject

state of poverty to prosperity. More so, a substantial capital injection into the economy will

raise income and saving to make the process of capital accumulation self-sustaining.



Vol.7, No.3, pp. 64-79, May 2019



Among the notable theories, Keynes (1936) in his hypothesis draws a link between public

expenditure and economic growth, implying that public sector expenditure is an exogenous

factor and a public instrument for increasing national income and concludes that increase in

government expenditure leads to higher economic growth. The link between the public

expenditure and growth not limited to the purview of Keynes. Romer (1986) developed a model

which revealed positive long run effect of government spending on economic growth while

investigating the impact of government spending on economic growth. In the study of Lucas

(1988), endogenous growth model was developed with human capital as the driver of perpetual

growth.

Barro (1990) and King and Rebelo (1990) endogenous growth models predict that government

spending and taxation will have both temporary and permanent effects on long run economic

growth. Also, Barro (1991) endogenous growth appeared to support empirical evidence

favouring the view that, a heavy government participation in economic activity tend to be

growth enhancing. Therefore, the introduction of endogenous growth models that incorporate

the government sector has led to the conclusion that fiscal policy can affect the long run growth

rate of an economy (Barro and Sala-i-Martin, 1992). This provides a sort of linkage between

government spending and economic growth. Harvey Leibenstein (1957) posited that

developing countries are generally characterized by vicious circle of poverty, which ranks them

around a low income per capita equilibrium state. The critical minimum effort required to raise

the per capita income to a level at which sustained development could be maintained justified

the government requirement effort for more spending that, will stimulate economic growth in

the developing countries.

A lot of research works have been carried out on the link between public expenditure and

economic growth yielding conflicting results. The conflicting remarks in this regard cut across

countries and economies have been left unresolved both in theoretical and empirical among

scholars. The inconclusive proofs from the existing literature further provided evidence for

more research study. However, the complexities of the size, structure and growth of public

sector expenditure have increased tremendously, especially in developing countries. In a study

carried out by Olugbenga and Owoeye (2007), on the relationships between government

spending and economic growth for a group of 30 developing countries, a long run relationship

existed between government spending and economic growth. Also, the result indicated a

unidirectional causality running from government spending to economic growth for 16 out of

the 30 countries investigated. In a survey of 102 studies on the economic effects of military

spending, Dunne and Uye (2010) observed that military expenditure has negative impact on

growth in 39% of cross-country and 35% of case studies. While 20% of the survey revealed a

positive relationship between military spending and economic growth.

Gisore et al (2014) investigated how government expenditure contributes to economic growth

in East Africa. Most existing studies examining the relationship between expenditure and

economic growth show conflicting results and mainly focus on aggregate expenditure. Hence

this study focused on disaggregated expenditure over the period from 1980 to 2010. The

objective of the study was to establish these expenditures that have effects on growth using

balanced panel fixed effect model. Employing LLC test, this study tested for panel unit root

and found that only GDP was stationary at level. The study found that expenditures on health

and defense were positive and statistically significant on growth. In contrast, education and

agriculture expenditure were insignificant. According to the investigation of Fan and Rao



Vol.7, No.3, pp. 64-79, May 2019



(2003) on the effect of different types of government expenditure on overall economic growth

across 43 developing countries between 1980 and 1998 using OLS method and found mixed

result. The rise in government spending on agriculture and health was particularly strong on

promoting economic growth in African countries. The results showed that, among all types of

government expenditures, agriculture, education, and defence contributed positively to GDP

growth in Asia. In Latin America, health spending had a positive growth-promoting effect.

Structural adjustment programs had a positive growth-promoting effect in Asia and Latin

America, but not in Africa.

Also, the research work of Davarajan, Swaroop and Zou (1993) employed panel data for14

developed countries between 1970 and 1990 with OLS method, 5-year moving average. The

study examined various functional types of expenditure (health, education, transport, and

others) as explanatory variables and found that health; transport and communication have

significant positive effectwhile education and defence have a negative effect on economic

growth. Bleaney Gemmel, and Kneller (2001)in their study on the effect of government

expenditure on GDP growth, using panels of annual and period-averaged data for

22Organizations for OECD countries during 1970-95.By employing OLS and GLS methods,

the study found that productive expenditures enhance growth, but non-productive spending

does not, in accordance with the predictions of Barro’s (1990) model. M'amanja and Morrissey

(2005) while investigating the effects of fiscal policy on growth in Kenya, categorized

government expenditure into productive and unproductive expenditures. However, contrary to

expectations, productive expenditure has a strong negative effect on growth, whilst

unproductive expenditure was found to be neutral to growth. Although in the long run,

government investment expenditure was found to be beneficial to growth. Studies on the

relationship between government expenditure and economic growth in Nigeria, is still plagued

with divergent conclusions. For example, Akpan (2005), Nurudeen and Usman (2010)

employed a disaggregated approach to determine the components of government expenditure

that enhances growth and those that do not. Their result arrived at a common standpoint i.e.

that there is no significant association between most components of government expenditure

and growth in Nigeria.

Sevitenyi (2012) on the relationship between government expenditure and growth employed

both a disaggregated and aggregated approach of method. The study finding shows a

unidirectional causality running from government expenditure to economic growth, meaning

that government expenditure promotes economic growth. Fajingbesi and Odusola (1999), and

Omoke (2009) made similar findings to that of Sevitenyi (2012). Meanwhile, Biswas and Ram

(1986) reported an insignificant effect of expenditure on defence as regards economic growth

in Nigeria. But by categorizing government's expenditure into sectors, Loto (2011) found that

expenditures on agriculture and education were negatively related to economic growth in the

short run whereas expenditures on security, transportation and communication were positively

related to economic growth (although statistically not significant).

The studies of Abu-Bader and Abu-Quarnon Israel and Syria (2003a), Haliciogglu (2005) for

Turkey, Govindaraju et al. (2010) for Malaysia, Wahab et al. (2011) for Nigeria, Kalam and

Aziz(2009), for Bangladesh, Kumar (2009), for China, Hong-Kong, Japan, Taiwan and South

Korea, Keho, (2015), for Gabon, Senegal and Burkina Faso, Ebaidalla (2013), for Sudan and

Gisore et al (2014) for East Africa. All argued in favour of government spending as an

accelerator of economic growth. In the other way, studies against the support of relationship



Vol.7, No.3, pp. 64-79, May 2019



between government spending and economic growth include; Huang (2006), for China and

Taiwan, Magazzino (2010), for EU-countries, Dogan and Tang(2006), for Indonesia, Malaysia,

Singapore and Thailand, Abu-Bader and Abu-Quarn (2003), for Egypt and Chimobi (2009),

for Nigeria. Similarly, Gwartneyet al., (1998), Schaltegger and Torgler (2006), Mitchell (2005)

and others proved empirically against government spending as an impetus for economic

growth. However, Frimpong and Oteng-Abaiye (2009) found neither support for Wagner law

nor Keynesian view for three ECOWAS countries in the eco-zone sub-group called WAMZ

that is Gambia, Ghana and Nigeria. Their results suggest that decreasing or increasing

government spending might not be a necessary policy action to achieve the steady growth in

those economies understudied. Olulu, Erhietovwe and Adrew (2014) noted that, not only has

recent political developments engendered expenditure growth, the question on raising

additional and identifying alternative sources of revenue to supplement the rising needs of

governance have made it more imperative to take a more focused look at government activities,

especially its expenditures.

METHODOLOGY

Data and Measurement

The selection of the sample period and countries are based on the availability of annual data,

ranging from 1990 to 2017. The selected African countries is classified by World Bank. Hence

this work makes use of a balanced panel data of 20 African countries(four from each sub-

region); Angola, Benin, Botswana, Cameroun, Central African Republic, Chad, Egypt,

Equatorial Guinea, Ethiopia, Ghana, Kenya, Mauritius, Morocco, Namibia, Nigeria, South

Africa, Sudan, Tanzania, Togo and Tunisia.

The study considered panel series data on real GDP per capita, defence expenditure, GFCF (as

a proxy for infrastructural expenditure), health expenditure and education expenditure obtained

from World Development Indicator (WDI) online database which was published by the World

Bank. The variables above are measured as follows. Real Gross domestic product (RGDP) is

measured in current US dollars by using current exchange rates of domestic currency against

the US dollar. The GDP figures are divided by total population of the country to get per capita

GDP.

Annual growth of gross fixed capital formation (GFCF) based on U.S dollar. This includes

plant, machinery, and equipment purchases; and the construction of roads, railways, and the

like, including schools, offices, hospitals, private residential dwellings, and commercial and

industrial buildings. Defence expenditure (DEXP) measured in U.S dollar, this is the military

expenditure (% of general government expenditure). This includes all current and capital

expenditures on the armed forces, including peacekeeping forces, defence ministries and other

government agencies engaged in defence projects. Health expenditure (HEXP), this is the

general government expenditure on education (current, capital, and transfers), is expressed as

a percentage of total general government expenditure on all sectors (including health,

education, social services, etc.). It includes expenditure funded by transfers from international

sources to government. General government usually refers to local, regional and central

governments.

Model specification

Given that the goal is to investigate the long-run and short-run association between economic

growth and disaggregated public expenditure, the empirical analysis makes use of panel co-



Vol.7, No.3, pp. 64-79, May 2019



integration and panel vector error correction methodologies. To this end, the empirical analysis

employs a panel co-integration approach, as well as panel VECM tests to identify the long-run

and short-run relationships among the variables. Building on the work of Gisore et al (2014),

the study adopts the model stated below.

RGDPG = f (Open, Tot, Pop, Tg, Hea, Edu, Def, Agr,)……………………………..(1)

Where: RGDPG = Real Gross Domestic Product Growth, OPEN = Openness, TOT = Terms

of trade POP = Population, Tg = Total government expenditure, Hea = Health Expenditure,

Edu = Education Expenditure, Def = Defence Expenditure, Agr = Agricultural expenditure.

Based on the study objectives, the model is re-specified and employed a panel co-integration

and panel VECM techniques to analyse data in selected African countries. The panel linear

function of the model is thus; 𝑅𝐺𝐷𝑃𝑖𝑡 = 𝑓(𝐺𝐹𝐶𝐹𝑖𝑡 , 𝐷𝐸𝑋𝑃𝑖𝑡 , 𝐻𝐸𝑋𝑃𝑖𝑡 , 𝐸𝐷𝐸𝑋𝑃 𝑖𝑡 , 𝑣𝑖)………………………………………………………..(2)

RGDP represents economic growth. Disaggregated government expenditure includes GFCF

(expenditure on infrastructure), DEXP (defence expenditure), HEXP (health expenditure), and

EDEXP (expenditure on education).our main variable of interest. 𝑣𝑖represents individual fixed country effects. Similarly, countries are indicated by the subscript i (i=1, ...... ,N), while t

represents the time period (t=1, .......,T).

RGDPit = αit +β1GFCFit +β2DEXP +β3HEXP + β4EDEXP+vi +𝜀𝑡+ µit ……………(3)

Cross Dependence (CD) and unit root tests

We first identify whether the given series are cross-sectional dependent. To this end, the

empirical analysis employs Pesaran's (2004) CD test. To select the correct type of unit root test,

we must first test for cross-sectional dependence for the variables and the co-integrating

equation. Thus, we employ the Lagrange Multiplier (LM) and bias-adjusted Lagrange

Multiplier tests developed by Breusch and Pagan (1980) and Pesaran, Ullah, and Yamagata

(2008), respectively. It is well known that when T is larger than N (T > N, as is the case in this

paper), LM and LMadj tests are favourable to the tests suggested by Frees (1995) and Pesaran

(2004). The LM test has a χ2 distribution with a cross-sectional independence null hypothesis.

It is based on the sum of squared coefficients of correlation among cross-sectional residuals

obtained through ordinary least squares (OLS). However, the LM test is biased when the group

mean is equal to zero and the individual mean is different from zero. Therefore, Pesaran et al.

(2008) corrected for bias by including variance and mean in the test statistic. In this way, they

obtained the bias-adjusted LM test, which has standard normal distribution.

Panel unit root tests

Since none of the panel unit root test is free from some statistical shortcomings in terms of size

and power properties, it is better for us to perform several unit root tests to infer an

overwhelming evidence to determine the order of integration of the variables. In this paper

three panel unit root tests: Levin, Linand Chu (LLC 2002), Im, Peasaran and Shin (IPS, 2003),

and Breitung and Das (2005) tests are applied.

The LLC test is based on the assumption that the persistence parameters 𝜌𝑖 are common across cross-sections so that 𝜌𝑖 = 𝜌 for all 𝑖, but this assumption is not true for several variables. The second and third tests assume cross-sectional independence. This assumption is likely to be

violated for the selected variables. It has been found by Banerjee et al. (2001) that these tests

have poor size properties and have a tendency to over-reject the null hypothesis of unit root if



Vol.7, No.3, pp. 64-79, May 2019



the assumption of cross-section independence is not satisfied. Pesaran (2007) and Choi (2006)

have derived other tests statistics to solve this problem.

Levin, Lin and Chu (LLC, 2002) considered the following regression equation:

∆𝑦𝑖𝑡 = 𝛼𝑦𝑖𝑡−1 + ∑ 𝛾𝑖𝑗

𝑝𝑖

𝑗=1

𝑦𝑖𝑡−𝑗 + 𝑋𝑖𝑡′ 𝛿 + 𝜀𝑖𝑡 … … … … … … … … … … … … … … … … … … … . (4)

where, ∆𝑦𝑖𝑡 = 𝑦𝑖𝑡 − 𝑦𝑖,𝑡−1, here the assumption is 𝛼 = 𝜌 − 1 i.e. 𝜌𝑖 = 𝜌 for all 𝑖 , but allow the lag order for the difference terms 𝜌𝑖to vary across cross-sections. Here the null hypothesis to be tested is 𝐻0: 𝛼 = 0against the alternative hypothesis𝐻0: 𝛼 < 0. The null hypothesis indicates that there is a unit root while the alternative hypothesis indicates that there is no unit

root. Im, Pesaran and Shin (IPS, 2003) proposed the test statistic using the following model:

∆𝑦𝑖𝑡 = 𝛼𝑦𝑖𝑡−1 + ∑ 𝛾𝑖𝑗

𝑝𝑖

𝑗=1

𝑦𝑖𝑡−𝑗 + 𝑋𝑖𝑡′ 𝛿 + 𝜀𝑖𝑡 … … … … … … … … … … … … … … … … … … … … . (5)

where, ∆𝑦𝑖𝑡 = 𝑦𝑖𝑡 − 𝑦𝑖,𝑡−1, 𝑦𝑖𝑡(𝑖 = 1,2, … … … … … , 𝑛; 𝑡 = 1,2 … … … … . 𝑇) is the series under investigation for country i over period t, pi is the number of lags in the ADF regression and

𝜀𝑖𝑡errors are assumed to be independently and normally distributed random variables for all 𝑖 and t with zero mean and finite heterogeneous variance 𝜎𝑖

2. Both 𝛼𝑖and 𝜌𝑖 in Eq. (5) are allowed to vary across the countries. The null hypothesis to be tested is that each series in the panel

contains a unit root, i.e. 𝐻0: 𝛼𝑖 = 0 ∀ 𝑖 against the alternative hypothesis that some of the individual series have unit root but not all.

𝐻1: {𝛼𝑖 = 0; 𝑓𝑜𝑟 𝑖𝛼𝑖 < 0; 𝑓𝑜𝑟 𝑎𝑡 𝑙𝑒𝑎𝑠𝑡 𝑜𝑛𝑒 𝑖

Breitung and Das (2005) showed that when individual-specific trends are included, the IPS test

can suffer from a loss of power due to bias correction. He proposes an alternative test unit root

which corrects for the loss of power and shows that it has greater power than the IPS test. The

null hypothesis of Breitung’s test is that the panel series exhibits non-stationary difference, and

the alternative hypothesis assumes that the panel series is stationary.

Heterogeneous panel cointegration

Granger (1981) showed that when the series becomes stationary only after being differenced

once (integrated of order one), they might have linear combinations that are stationary without

differencing. In the literature, such series are called co-integrated’’. If integration of order one

is implied, the next step is to use co-integration analysis in order to establish whether there

exists a long-run relationship among the set of the integrated variables in question. Earlier tests

of co-integration include the simple two-step test by Engle and Granger(1987) (EG). However,

the EG method suffers from a number of problems. Therefore, this study shall follow the

recently developed panel co-integration tests by Pedroni (2004) provide a technique that allows

for using panel data thereby overcoming the problem of small samples, in addition to allowing

for heterogeneity in the intercepts and slopes of the co-integrating equation. Pedroni’s method

includes a number of different statistics for the test of the null of no co-integration in

heterogeneous panels. A group of the tests are termed ‘‘within dimension’’ (panel tests) and

the other group as ‘‘between dimension’’ (group tests). The ‘‘within dimension’’ tests pool the

data across the ‘‘within dimension’’. It takes into account common time factors and allows for

heterogeneity across members. The ‘‘between dimension’’ tests allow for heterogeneity of

parameters across members, and are called ‘‘group mean co-integration statistics’’.

Seven of Pedroni’s tests are based on the estimated residuals from the following long-run

model:



Vol.7, No.3, pp. 64-79, May 2019



𝑦𝑖𝑡 = 𝛼𝑖 + ∑ 𝛽𝑗𝑖𝑥𝑗𝑖𝑡 + 𝜀𝑖𝑡 ,

𝑚

𝑗=1

… … … … … … … … … … … … … … … … … … … … … … … … … … … … … … … … . . (6)

where𝜀𝑖𝑡 = 𝜌𝑖𝜀𝑖(𝑡−1) + 𝑤𝑖𝑡 are the estimated residuals from the panel regression. The null

hypothesis tested is whether 𝜌𝑖 is unity. The seven statistics are normally distributed. The statistics can be compared to appropriate critical values, and if critical values are exceeded then

the null hypothesis of no cointegration is rejected implying that a long-run relationship between

the variables does exist.

Panel Vector Error Correction Model

In economics, deviations from a long-run equilibrium are possible, but these errors are

characterized by a mean revision back to its long-run equilibrium (Pfaff & Gentleman 2008

p.76). The question is how to model this dynamic behaviour. Engle & Granger (1987) that

proposed a two-step estimation technique to model dynamic behavior of I(1) variables that are

cointegrated, which is implemented in this paper. In the first step, the following model is

estimated:

RGDPit =αi +β1iEXPVit +eit

where i=1,2….,N is the number of countries in the panel, t=1,2,….,T is the number of time

periods, RGDP is the economic growth rate, EXPV is the expenditure variables and e is the

residuals. The residuals are obtained:

êit = RGDPi,t -α̂I,t- β̂1iEXPVi,t The lagged residual (êit-1) now contains information about the long-term relationship and the

adjustment process to its long run equilibrium (Asteriou and Hall 2011 p. 365). in this paper,

the lagged residual(êit-1) is represented by ECTit-1.The next step in Engle & Granger (1987)

two-step procedure is to estimate a system of equations where the error correction term is

incorporated with the short dynamics (Hill et al. 2011 pp.499-502). The system is written as:

∆RGDPit= α11i +β11i∆EXPVit+∑ ∅𝑗1𝑖2𝑘=1𝑗=1

∆EXPVi,t-k + 𝜆11𝑖ECTi,t-1+𝜇11𝑖,𝑡

∆EXPVit= α22i +β22i∆RGDPit+∑ 𝛾𝑗2𝑖2𝑘=1𝑗=1

∆RGDPi,t-k + 𝜆12𝑖ECTi,t-1+𝜇22𝑖,𝑡

Where ∆ is the first difference operator, k is the lag length, α, β,∅, 𝜆, 𝑎𝑛𝑑 𝛾are slope coefficient j in equation and 𝜇 is the residuals.

RESULTS AND DISCUSSION OF FINDINGS

Cross Sectional Dependence and Unit Root Test

As seen in table 1 below, all the LM tests including Pesaran CD reveal the existence of cross-

sectional dependence at 1% significance level for all the variables. Hence we conduct a unit

root test which allow for cross-sectional dependence. Table 2 is the panel unit root test result

which shows evidenced of first order integration as all the variables are stationary at first

difference I(1). This implies that we move on to co-integration test as the order of integration

informed the use of co-integration.



Vol.7, No.3, pp. 64-79, May 2019



Table 1: Cross Sectional Dependence Test

Variables Breusch-Pagan LM Pesaran scaled LM Bias-corrected scaled LM

Pesaran CD

Rgdp 355.9476(0.0000) 8.51293(0.0000) 8.142562(0.0000) 9.9611(0.0000)

Gfcf 289.3434(0.0000) 5.09621(0.0000) 4.725839(0.0000) 1.84099(0.0000)

Dexp 1443.834(0.0000) 64.3203(0.0000) 63.94996(0.0000) 10.9208(0.0000)

Hexp 2401.936(0.0000) 113.470(0.0000) 113.0995(0.0000) 2.2539(0.0242)

Edexp 289.3434(0.0000) 5.09621(0.0000) 4.72584(0.0000) 1.84099(0.0000)

Notes: P-values of the test statistics are presented in parentheses

Source: Authors Computations

Table 2: Panel Unit Root Test (LLC Breit IPS)

Variables level I(0) Difference I(1)

LLC Breit IPS LLC breit IPS

Rgdp -0.21491 -3.65822 -3.73868 -

10.4645*** -

9.52295*** -17.361***

(0.4149)

( 0.0001) (0.0001) (0.0000) (0.0000) (0.0000)

Gfcf -1.62935 -0.82239 -5.74536 -

15.3215*** -

11.0429*** -

20.9873***

0.0516 0.2054 0.0000 (0.0000) (0.0000) (0.0000)

Dexp -0.34562 -3.85814 -4.55632 -

6.37384*** -

5.02807*** -12.550***

0.3648 0.0001 0.0000 (0.0000) (0.0000) (0.0000)

Hexp -0.83286 -1.80243 -3.76883 -

9.78705*** -

10.0036*** -

12.4679***

0.2025 0.0357 0.0001 (0.0000) (0.0000) (0.0000)

Edexp -1.7024 -0.59419 -5.49155 -

9.32313*** -

13.2122*** -

13.8179***

0.0443 0.2762 0.0000 (0.0000) (0.0000) (0.0000)

*Significant at 10%, **Significant at 5%, and ***Significant at 1%.

The asterisks indicate the rejection of null hypothesis of unit root


Results of Panel Cointegration Test

The hypothesis of cointegration between all variables is tested using pedroni (2004)

cointegration tests. As seen in table 3 below, all the three assumptions (no trend, trend and

intercept and no trend or intercept) indicate the presence of cointegration among the variables.

Thus majority of between and within dimension statistics indicate that the null hypothesis of

no co-integration is rejected at 1% and 5% significance levels. This empirical finding further

proves the presence of long run equilibrium relationship between economic growth and

expenditure variables.



Vol.7, No.3, pp. 64-79, May 2019



1. Table 3: Panel Co-integration Test Result

Within-dimension No trend intercept and trend none

Stat Prob w-stat prob Stat Prob w-stat prob stat prob w-stat Prob

Panel v-Statistic -1.977

0.9760

-2.155 0.9844

-4.181 1.0000

-4.269 1.0000

-0.739 0.7700

-1.008 0.8432

Panel rho-Statistic

-3.198***

0.0007

-3.267***

0.0005

-1.396* 0.0814

-1.422* 0.0775

-4.285***

0.0000

-4.268***

0.0000

Panel PP-Statistic

-9.593***

0.0000

-10.837***

0.0000

-9.412***

0.0000

-11.731***

0.0000

-9.923***

0.0000

-10.580***

0.0000

Panel ADF-Statistic

-6.921***

0.0000

-9.037***

0.0000

-5.854***

0.0000

-8.992***

0.0000

-8.162***

0.0000

-9.794***

0.0000

between-dimension

Stat Prob Stat Prob stat prob

Group rho-Statistic

-3.01***

0.0013

-1.135

0.1282

-4.371***

0.0000

Group PP-Statistic

-14.2***

0.0000

-16.521***

0.0000

-14.771***

0.0000

Notes: Null hypothesis: No cointegration, lag selection: Automatic AIC with a max lag of 5. ***designate the significance at 1% significance level, **designate the significance at 5% significance level while *designate the significance at 10% significance level. Source: Authors Computations

DISCUSSION OF RESULTS OF PANEL VECM

In this paper, we estimated the long-run and short-run parameters using panel VECM. Hence,

the long-run estimates indicate that expenditure on infrastructure (gfcf), health expenditure

(hexp) and expenditure on education (edexp) have positive impact on economic growth (rgdp)

of the selected African countries at about 2%, 6% and 2% respectively, but only expenditure

on infrastructure is significant. This result implies that expenditures on infrastructure, health

and education recorded only a meagre contribution to economic growth in African countries

investigated. This finding corroborates the study of Olugbenga and Owoeye, (2007). Also,

defence expenditure (dexp) has both negative and insignificant impact on economic growth in

the long-run. This is in line with the findings of Dunne and Uye (2010).

However, the short-run estimates revealed that expenditure on infrastructure at lag 1 and 2 have

indirect impact on economic growth in both lags. Meanwhile, a unit rise in expenditure on

infrastructure contributed about 3% significant reduction in economic growth at lag 1. Defence

expenditures and education expenditures at both lags have indirect and insignificant influence

on economic growth while health expenditure has direct and insignificant impact on economic

growth at all lags. Most importantly, the speed of adjustment is high as ECT shows negative

and significant implying that about 72% of the error is corrected annually from the short run to

long run. This further indicates that the speed of adjustment is very fast in correcting the error

from short run to long run. Finally, to examine if there is short run equilibrium, we estimated



Vol.7, No.3, pp. 64-79, May 2019



Wald test which indicates acceptance of null hypothesis. Thus p-value is insignificant and we

accept the null hypothesis and conclude that there is no short run equilibrium relationship

among the variables.

Table 4: Panel VECM of Long run and short run

PANEL VECM

LONG RUN REGRESSION SHORT RUN ESTIMATES

Variables Coef Std errors t-stat* variables

GFCF(-1) -0.02867 (0.01396) [-2.05361] Coef prob t-stat*

DEXP(-1) 0.004854 (0.04888) [ 0.09932] GFCF(-1) -0.03302 0.0125 -2.50792

HEXP(-1) -0.06974 (0.19957) [-0.34945] GFCF(-2) -0.00574 0.6601 -0.43999

EDEXP(-1) -0.02868 (0.02878) [-0.99634] DEXP(-1) -0.04618 0.534 -0.62238

DEXP(-2) -0.02048 0.7762 -0.28449

HEXP(-1) 0.729968 0.265 1.115998

HEXP(-2) 0.441123 0.5157 0.650475

EDEXP(-1) -0.02214

0.4597 -0.73996

wald test EDEXP(-2) -0.02329

0.4255 -0.79762

Stat 9.60016 ECT -0.72108 0.0000

Prob 0.2942 R² 0.420203


The relation to existing literature

The positive impact of health expenditure and defence expenditure on growth is in line with

the works of Fan and Rao (2003) and Devarajan, Swaroop and Zou (1993). The negative impact

of defence expenditure on economic growth is in line with the work of Devarajan, Swaroop

and Zou (1993) and Fan and Rao (2003), but against the work of Loto (2010). Also the positive

impact of education expenditure in the long run is in line with the wok of Fan and Rao (2003)

and Loto (2010) but against the work of Devarajan, Swaroop and Zou (1993).

Implications to Research and Practice

The main objective of this study is to investigate the long-run and short-run equilibrium

relationship between economic growth and disaggregated public expenditure in Africa with

panel data spanning 1990 through 2017. To achieve this aim, the present paper employed panel

co-integration based on Pedroni (2004) and a PVECM estimated with Engle & Granger´s

(1987) procedure. The result provided useful evidence of co-integration between economic

growth and various types of government expenditure. The panel VECM result indicates that

there is a long run equilibrium association between economic growth and various government

expenditures as indicated by the error correction term which was high, rightly signed and

significant. In the long run also, only infrastructural expenditures impacted positively and

significantly on economic growth in Africa. Education and health have positive but

insignificant effect on economic growth. On the short run only expenditure on infrastructure is

significant in influencing economic growth in Africa although, it is inversely related to

economic growth. The wald test revealed that there is no short run equilibrium relationship

meaning that equilibrium only occur in the long run. This points to the fact that expenditure is

a long run issue as the evidence in real life can be seen over a period of time. The study



Vol.7, No.3, pp. 64-79, May 2019



established that massive investment in infrastructural development will impact positively

economic growth in selected African Countries. The studies unravel that expenditure on health

sector and education has not met up with the necessary requirement to influence economic

growth which are major pivotal to the selected African economies. Also, the research works

found out that the impact of gross government expenditure over the study period has no

significant contribution to economic growth which could be subject to gross mismanagement

and corruption which had retarded the economy performance over the years.

CONCLUSION

The study concluded by recommending that the selected African countries policy makers to

focus on developing their health, infrastructure and education sectors which has not contributed

significantly enough to economic growth. This will enhance human capital formation which

will ultimately promote economic growth. The study affirmed that despite huge funding on

defense, the insecurity is still persistent in some of the selected African countries, yet these

have not contributed to economic growth. Hence, the study recommends stiffer penalty on

economic managers in order to reduce corruption and mismanagement of government funds.

Finally, policy maker should put in place check and balance measures to raise the need for

transparency, probity and accountability on how public expenditure is spent.

Future Research

Further researches should improve on comprehensive analysis of defense spending that will be

used to enhance the representativeness and quality of the result of the selected African

Countries.

References

Abu-Bader, S., and Abu-Qarn, A. S. (2003). Government expenditures, military spending and

economic growth: Causality evidence from Egypt, Israel, and Syria. Journal of Policy

Modeling. 25(6-7), 567-583.

Akpan, N.I. (2005). Government Expenditure and Economic Growth in Nigeria: A

Disaggregated Approach. CBN Economic and Financial Review, 43(1).

Asteriou, D. and Hall, S.G. (2011). Applied Econometrics. New York, NY: Palgrave

Macmillian.

Banerjee, A., Massimiliano, M. and Chiara, O. (2001). Some cautions on the use of panel

methods for integrated series of macroeconomics data. European University institute,

Discussion paper 00/20.

Barro R, Sala-i-Martin X, (1992). Public Finance in Models of Economic Growth. Review of

Economic Studies. 59: 645-661.

Barro, R. (1990). Government Spending in a Simple Model of Endogenous Growth. Journal

of Political Economy. 98(5):103-125.

Barro, R. J., (1991). Economic Growth in a Cross Section of Countries. Quarterly Journal

Economics. 07-444

Barro, R. J., and Sala-i-Martin, X. (1992). Public Finance in Models of Economic Growth.

Review of Economic Studies. 9, 645-661.

Biwas B and Ram R. (1986). Military Expenditures and Economics Growth in Less Developed

Countries: An Augmented Model and Further Evidence. Econ. Dev. Cultural Change.

No.34: 362-336.



Vol.7, No.3, pp. 64-79, May 2019



Bleaney, M., Gemmell, N. and Kneller, R., (2001). Testing the endogenous growth model:

public expenditure, taxation and growth over the long run. Canadian Journal of

Economics/Revuecanadienne d’économique, 34(1), pp.36–57.

Breitung, J. (2002). Nonparametric tests for unit roots and co-integration. Journal of

econometrics. 108 (2), 343-363.

Breitung, J. and Das, S. (2005). Panel unit root tests under cross‐sectional dependence. Statistica Neerlandica. 59(4), 414-433.

Breusch, T. S., and Pagan, A. R. (1980). The Lagrange multiplier test and its applications to

model specification in econometrics. The review of economic studies, 47(1), 239-253.

Chimobi, O.P. (2009). Government expenditure and national income. A causality test for

Nigeria. European Journal of Economic and political Studies. 2; 1-11.

Choi, I. (2006). Non-stationary panelsIn palgrave Handbook of Econometrics. Vol 1, Mills TC,

Patterson K. (eds). Palgrave Macmillan: Basingstoke; 511-539.

Dar Atul A, AmirKhalkhali S, (2002). Government Size, Factor Accumulation and Economic

Growth: Evidence from OECD Countries. Journal of Policy Modeling. 24(7/8): 679-692.

Davarajan, S., Swaroop, V. and Zou, H. (1993). What Do Governments Buy? The Composition

of Public Spending and Economic Performance. Policy Research Working Paper.

WPS1082.

Devarajan, S. Swaroop, V. and Zou, H. (1996). The Composition of Publiuc Expenditure and

Economic Growth. Journal of Monetary Economics. 37: 313-344.

Dogan, E. and Tang, T. C. (2006). Government Expenditure and National Income:

causalityTest For Five South East Asian Countries, International Business & Economics

Research Journal, 5(10): 49-58.

Domar, E. D. (1946). Capital Expansion, Rate of Growth and Employment. Econometrica. 14

(2): 137 – 147.

Domar, E. D. (1946). Capital expansion, rate of growth, and employment. Econometrica,

Journal of the Econometric Society. 137-147.

Dunne, J. P. and Uye, M. (2010). Military Spending and Development. In Global Arms Trade,

edited by A. Tan, 293–305. London: Routledge.

Ebaidalla, E. M. (2013). Causality between government expenditure and national income:

Evidence from Sudan. Journal of Economic Cooperation and Development. 34(4): 61-

76.

Ebaidalla, E. M. (2013). Causality between government expenditure and national income:

Evidence from Sudan. Journal of Economic Cooperation and Development. 34(4), 61-

76.

Engle, R. F. and Granger C.W.J. (1987). Cointegration and error correction: representation,

Estimation and Testing. Econometrica, 55: 251-276.

Engle, R.F. and Granger, C.W.J (1987). Co-integration and error correction; Representation,

estimation and testing. Econometrica. 55; 251-276

Fajingbesi, A. A. and Odusola, A. F. (1999). Public Expenditure and Growth. A Paper

Presented at a Training Programme on Fiscal Policy Planning Management in Nigeria,

Organized by NCEMA, Ibadan, Oyo State, 137-179.

Fan, S. and Rao, N. (2003). Public Spending in Developing Countries: Trend, Determination

and Impact. EPTD Discussion Paper No. 99, Washington, D.C.

Frees, E. W. (1995). Assessing cross-sectional correlation in panel data. Journal of

econometrics. 69(2), 393-414.



Vol.7, No.3, pp. 64-79, May 2019



Frimpong, J. M., and Oteng-Abayie, E. F. (2009). Does the Wagners Hypothesis Matter in

Developing Economies? Evidence from three West African Monetary Zone (WAMZ)

Countries. American Journal of Economics and Business Administration, 1(2), 141-147.

Gisore, N., Kiprop, S., Kalio, A., Ochieng, J. and Kibet, L. (2014). Effect of Government

Expenditure on Economic Growth in East Africa: A Disaggregated Model. European

Journal of Business and Social Sciences. 3 (8): 289-304

Govindaraju, D. R., Houle, D., and Omholt, S. (2010). Phenomics: the next challenge. Nature

reviews genetics. 11 (12), 855.

Granger, C. W. (1981). Some properties of time series data and their use in econometric model

specification. Journal of econometrics. 16(1), 121-130.

Gwartney, J. D., Lawson, R. A. and Holcombe, R. G. (1998). The size and functions of

government and economic growth. Joint economic committee. Washing Grossman, P.

(1988). Growth in government and economic growth: The Australians experience.

Australians Economic papers. 27: 33-45

Halicioglu, F. (2005). An ARDL model of aggregate tourism demand for Turkey (No.

0503005). University Library of Munich, Germany.

Harrod, R. F. (1960). Second essay in dynamic theory. The Economic Journal, 277-293.

Harrod, R.F. (1960). The Theory of Economic Growth. The Economic Journal. 70, (278): 277-

293.

Huang, C.J. (2006). Government expenditure in China and Taiwan: Do they follow Wagners’s

Law? Journal of Economic Development. 31(2), 139-148

Im, K. S., Pesaran, M. H., and Shin, Y. (2003). Testing for unit roots in heterogeneous

panels. Journal of econometrics. 115(1), 53-74.

Iyoha, M. and D. Oriakhi. (2002). Explaining African Economic Growth Performance: The

case of Nigeria. Revised Interim Report on Nigerian Case Study, Prepared for the African

Economic Research Consortium Research Project, Explaining African Economic Growth

Performance

Kalam, M. A., & Aziz, N. (2009). Growth of government expenditure in Bangladesh: An

Empirical enquiry into the validity of Wagner's law. Global Economy Journal. 9(2).

Keho, Y. (2015). Foreign Direct Investment, Exports and Economic Growth: Some African

Evidence. Journal of Applied Economics and Business Research. 5(4).

Keynes, J. M. (1936). The General Theory of Employment, Interest and Money, published by

Harcourt, Brace and Company, and printed in the U.S.A. by the Polygraphic Company

of America, New York; First Published: Macmillan Cambridge University Press

King, R. G., and Rebelo, S. (1990). Public policy and economic growth: Developing

neoclassical implications. Journal of Political Economy. 98(5 & 2), S126-S150.

Kumar, S. (2009). An empirical evaluation of export demand in China. Journal of Chinese

Economic and Foreign Trade Studies, 2(2), 100-109.

Leibenstein, H. (1957), Economic Backwardness and Economic Growth. New York: Wiley.

Levin, A., Lin, C. F. and Chu, J. (2002). Unit Root Tests in Panel Data: Asymptotic and Finite

Sample Properties. Journal of Econometrics. 108:1–24.

Loto, M. A. (2011). Impact of government sectoral expenditure on economic growth. Journal

of Economics and international Finance. 3(11), 646-652.

Lucas, R. E. (1988). On the mechanics of economic development. Journal of monetary

economics. 22(1), 3-42.

M’Amanja M. and Morrissey O. (2005). Fiscal Policy and Economic Growth in Kenya.

CREDIT Research Paper No 05/06.



Vol.7, No.3, pp. 64-79, May 2019



Magazzino, C. (2010). Wagner’s Law in Italy: Empirical Evidence from 1960 to 2008.Global

and Local Economic Review. 14 (1); 91-116.

Mitchell, J. D. (2005). The Impact of Government Spending on Economic Growth,

Backgrounder, 1831. [www.heritage.org/research/budget/bg1831.cfm]

Niloy, B., Emranul, H.M., and Osborn, D.R. (2003). Public Expenditure and Economic

Growth: A Disaggregated Analysis for Developing Countries.

[www.socialsciences.manchester.ac.uk/cgbcr/dpcgbr.30.pdf]

Niloy, B., Emranul. M.H and Denise. R.O (2003): Public Expenditure and Economic Growth:

A Disaggregated Analysis for Developing Countries, JEL, Publication.

Nurudeen, A. and Usman, A. (2010). Government Expenditure and Economic Growth In

Nigeria, 1970-2008: A Disaggregated Analysis. Business and Economics Journal,

Volume 2010: BEJ-4. http://astonjournals.com/bej.

Olugbenga, A. O. and Owoye, O. (2007). Public expenditure and economic growth: New

evidence from OECD countries. Business and Economic Journal. 4(17), 13-25.

Olulu, R. M., Erhietovwe, E. K. and Adrew, U. (2014). Government Expenditures and

Economic Growth: The Nigerian Experience. Mediterranean Journal of Social Sciences.

5(10), 89-94.

Olulu, R. M., Erhieyovwe, E. K., and Andrew, U. (2014). Government expenditures and

economic growth: The Nigerian experience. Mediterranean Journal of social

sciences, 5(10), 89.

Omoke P. (2009). Government Expenditure and National Income: A Causality Test for Nigeria.

European Journal of Economics and Political Studies. 2:1011.

Oteng-Abayie E.F (2011). Government Expenditure and Economic Growth in Five ECOWAS

Countries: A Panel Econometric Estimation. Journal of Economic Theory. 1: 1141.

Oteng-Abayie, E. F. (2011). Government Expenditure and Economic Growth in Five

ECOWAS Countries: A Panel E Econometric Estimation. Journal of Economic Theory.

5(1), 11-14

Peacock, A.T., and Wiseman, J. (1961). The Growth of Public Expenditure in the United

Kingdom. Princeton University Press.

Pedroni, P. (2004). Panel Co-integration: Asymptotic and finite sample properties of pooled

time series tests with an application to the PPP hypothesis. Econ. Theory. 20; 597-625.

Pesaran, M. H. (2004). General diagnostic tests for cross section dependence in panels. CESifo

Working Paper Series No. 1229; IZA Discussion Paper No. 1240. Available at

SSRN: https://ssrn.com/abstract=572504

Pesaran, M. H., A. Ullah and T. Yamagata (2008). A bias-adjusted LM test of error cross

section independence. Econometrics Journal 11, 105–127.

Pesaran, M. H., Ullah, A., & Yamagata, T. (2008). A bias‐adjusted LM test of error cross‐section independence. The Econometrics Journal. 11(1), 105-127.

Pesaran, M.H. (2007). A simple panel unit root test in the presence of cross section dependence.

Journal of Applied Econometrics. 22: 265-312.

Pfaff, B. and Gentleman, R. (2008). Analysis of integrated and co-integrated time series with

R. 2 ed. New York : Springer.

Romer, P. M. (1986). Increasing returns and long-run growth. Journal of political economy.

94(5), 1002-1037.

Rostow, W. W. (1960). The stages of growth: A non-communist manifesto. Cambridge:

Cambridge University Press.

Roux A (1994) Defense, Human Capital and Economic Development in South Africa. African

Defense review no. 19.

http://www.eajournals.org/http://astonjournals.com/bejhttps://ssrn.com/abstract=572504


Vol.7, No.3, pp. 64-79, May 2019



Schalteger, C. A. and Torgler, B. (2006). Growth Effects of Public Expenditure on the state

and local level: Evidence from a sample of rich government. Applied Economics, 38,

1181- 1192.

Sevitenyi, I. N. (2012), Government Expenditure and Economic Growth in Nigeria: An

Empirical Investigation (1961-2009). The Journal of Economic Analysis, 3(1), pp38- 51.

Solow, R. M. (1956). A Contribution to the Theory of Economic Growth. Quarterly Journal

of Economics, 70(1): 65-94.

Wagner, A. (1893). Grundlegung der politischen Okonomie, 3rd ed.,Leipzig.

Wahab, A. L. and Rihanat, I.A. (2011). An analysis of government spending on educational

sector and its contribution to GDP in Nigeria. International journal of Financial

Economics and Econometrics. 3 (1): 163-170

Date post:	17-Feb-2021
Category:	Documents
Upload:	others
View:	2 times
Download:	0 times

IMPACT OF DISAGGREGATED PUBLIC EXPENDITURE ......Harvey Leibenstein (1957) posited that developing...

Documents